The goal of this collaborative R01 proposal (a coordinating proposal is being submitted by Dr. Richard M. Myers and colleagues at HudsonAlpha Institute) is to use high-throughput DNA sequencing and genotyping technologies to identify genes and pathways that contribute to the risk for bipolar disorder (BD). This proposal builds on the active collaboration between our groups at the University of Michigan and HudsonAlpha on BD and other mood disorders. Our research team combines strengths in high-throughput genetics and genomics and development and application of innovative computational and statistical methods to maximize the benefits of these new technologies. In Specific Aim 1, we will sequence DNA from 2,000 individuals, 1,025 with BD and 975 controls, at >4X coverage for the genome and at >60X average coverage for the 44 Mb exome. In Specific Aim 2, we will carry out BD association analyses based on sequence data from these 2,000 samples and imputed data from at least an additional 6,085 BD case and 7,116 control GWAS samples to identify BD- associated variants in at least 7,110 cases and 8,091 controls. We will analyze variants with MAF>0.5% individually. For variants with MAF<1%, we will use burden tests designed to identify regions where clusters of rare variants are more common in cases than controls (or vice versa). For imputation, we will use sequence data from this project, the 1000 Genomes Project, and our GoT2D project as reference sets, and BD case- control sample individuals with GWAS genotype data from previously-published GWAS as target sets. In Specific Aim 3, we will use custom SNP arrays to genotype ~5,000 sequence variants and deep sequencing to resequence 500 selected genes in 7,592 individuals, 3,854 with BD and 3,738 controls, and carry out BD association analysis on the resulting data. This set will include the 2,000 individuals sequenced in Specific Aim 1 to validate the sequence-based genotypes and 5,592 additional individuals to validate imputation-based association findings and/or strengthen evidence for rarer variant association. In collaboration with Dr. Pamela Sklar, we will follow-up the most interesting SNPs by genotyping 32,000 BD cases and 30,000 controls and carry out a meta-analysis of the resulting data. In Specific Aim 4, we will share data and methods to support similar studies for BD and other psychiatric phenotypes, and more broadly across the scientific community. Completion of these aims will provide new insights into disease mechanism that have the potential to catalyze breakthroughs in BD prevention, treatment, and diagnosis.